Flaky under parallel load: MeetingUrl_AcceptsTheStorageLimitAndRejectsOverflow

[peterdrier]

Humans.Workgroups.Tests.Services.WorkgroupInputValidationTests.MeetingUrl_AcceptsTheStorageLimitAndRejectsOverflow(editing: True) intermittently times out during the full-suite dotnet test run in CI.

Run: https://github.com/peterdrier/Humans/actions/runs/35544429622 (job 106167805200)

Assertion:

[Test Cleanup Failure (MeetingUrl_AcceptsTheStorageLimitAndRejectsOverflow(editing: True))]: System.OperationCanceledException : The operation was canceled.
Test execution timed out after 5000 milliseconds

Stack trace bottoms out in EF Core InMemory (InMemoryShapedQueryCompilingExpressionVisitor.QueryingEnumerable.MoveNextAsync) via WorkgroupRepository.GetWorkgroupAsync → WorkgroupService.RequireAsync → CreateMeetingAsync.

Observations:

• Seen on PR Issue queues: each section declares its own, Issues stops listing them #1766, whose diff over the last-green commit is docs-only (Humans.Issues/Docs/Issues.md) — Workgroups is untouched by that PR.
• Passes 2/2 when run in isolation (dotnet test tests/Humans.Workgroups.Tests --filter FullyQualifiedName~MeetingUrl_AcceptsTheStorageLimitAndRejectsOverflow).
• Consistent with other reports of intermittent parallel-load timeouts on this branch, e.g. Shifts SelfSignup_NonActiveAccount_DoesNotWriteOrNotify and Workgroups Invalid_reasons_render_queue_with_original_text_and_error, both also green in isolation.

Likely a shared-resource contention / 5s timeout budget issue when many test assemblies run in parallel, not a real product bug.